Sweet Spot on Zoom Lenses on APS-C CMOS sensors?

Question

I have made the conscious decision not to go full-frame on my camera body purchase but to spend the money on top quality 'L' lenses for my Canon 80D. More explicitly, I have the 24-105mm f/4 L and 70-200mm f/4 L and do not require a low light capability to warrant the price increase on going down to 1.8 etc. I spend most of my time out in the Scottish Highlands and although there are some scenarios where I do require more control of my DOF, invariable I am disappointed by my usual landscape photos at F11. Why is this? Can anyone recommend what f-stop I should be using for a 'generic' landscape shot in order to get it pin sharp?

Answer 1

First, let's talk about diffraction

Your 80D has a 26 MP sensor with pixels only 3.72µm wide. That figures out to a diffraction limited aperture (DLA) of f/6.0.

Diffraction Limited Aperture (DLA) is only applicable at 100% viewing size. This is because DLA assumes a Circle of Confusion (CoC) equal to the pixel pitch of a particular sensor. The effects of diffraction at the DLA are only observable if the resulting image is magnified enough that the viewer can discretely resolve individual pixels. For a 26MP image viewed on a 23" HD (1920x1080) monitor that is the equivalent magnification of a 60"x40" print!

Diffraction at the DLA is barely visible when viewed at 100% (1 image pixel = 1 monitor pixel) on a display. As sensor pixel density increases, each pixel gets smaller and the DLA gets wider. DLA does not mean that narrower apertures should not be used. It is where image sharpness begins to be compromised for increased DOF. Higher resolution sensors generally continue to deliver more detail well beyond the DLA than lower resolution sensors until the "Diffraction Cutoff Frequency" is reached (a much narrower aperture). The progression from sharp to soft is not an abrupt one. Current Canon DLSRs may have a DLA as low as f/5.2 (90D) and as high as f/10.6 (EOS 1D X Mark III). Most other manufacturers' DSLR offerings fall somewhere along the same lines.

Ultimately you must consider all of the factors involved to decide what is the best aperture to use for a particular photograph. Many times, it will be a compromise between several factors such as more depth of field (narrow aperture) and usable shutter speed and ISO (wide aperture).

For more, please see this answer to Do smaller apertures provide more depth of field past the diffraction limit, even if peak sharpness suffers?

Of course, we're assuming you're already using the camera on a stable tripod with a remote release and using mirror lockup when shutter duration is longer than about 1/200 second.

For more about other things that might cause blurry images, please see How do I diagnose the source of focus problem in a camera?

Now let's talk about gear and skill

I'm a firm believer in using the gear you've got until you can articulate exactly where your gear is holding you back from being able to get a shot you are capable of getting. A lot of folks who are dissatisfied with the results of their efforts will go chasing after "better", more expensive gear thinking it will make them a better photographer.

It won't.

What they do not realize is that the dissatisfaction they have with their work has very little to nothing to do with the limits of their gear and almost everything to do with their limits as a photographer.

From my answer to Does the camera matter? :

While it is true that better gear won't make you a better photographer, it is equally true that any photographer is limited by the capabilities of the gear being used. It's not just "lesser" types of gear that technically constrain photographers. Even the very best available photographic gear imposes technical limits on what may be done.

There's an old saying that has been around photography for a very long time:

Gear doesn't matter.

It's certainly true, but it is only half the truth. The rest of the truth is this:

Gear doesn't matter - until it does.

When the technical capabilities of your gear are not up to the task for the shots you want to capture, then and only then will the gear matter.

When your gear does matter, you'll know. It will matter because the gear you are using will limit you from doing work that you wish to do and that you have the skill and knowledge to pull off. Until you reach that point, the gear you are currently using is perfectly fine for you.

With that in mind, let's discuss what you are trying to accomplish and what might be the best way to get there.

You express the desire to get "pin sharp" landscape photos.¹ You also say that you have consciously decided to use an APS-C camera in order to allow you to spend more on higher quality lenses.

Those decisions and the particular "L" lenses you chose may have been a bit misguided. Here are a few reasons why:

• For not much more than the price of a new 80D, one can get a pretty good full frame camera, such as (staying in the Canon ecosystem) the 6D Mark II. The 80D was introduced at a price of $1,199 in the U.S. and currently sells for a list price of $1,099 less a $200 "instant rebate" that makes the net cost $899. The 6D Mark II was introduced at $1,999 in the U.S., but is currently selling for a list price of $1,599 less a $400 "instant rebate" for a net price of $1,199. The current $300 difference in price between the 80D and 6D Mark II are well less than the cost of a single premium lens. Both options, APS-C and full frame, have advantages and disadvantages when compared to the other. But for the way most landscape photographers shoot, a FF camera is usually the better overall choice.
• The same lens will not perform equally on different cameras. This is particularly the case when the sensors for each camera are different sizes. Used with a larger sensor camera, the same lens can perform better due to the smaller enlargement ratio needed to get to the same display size. To view at 8x12 inches, an image from a Canon APS-C camera must be enlarged by a factor of 13.5X. To view at 8x12 inches, an image from a full frame camera must be enlarged by a factor of 8.45X. When you enlarge an image, you also enlarge the blur in that image. The more you enlarge, the more noticeable the blur becomes. Some blur that is still not perceivably different from a point by the viewer at 8.45X enlargement will be noticeably burry by the viewer when enlarged by a factor of 13.5X. If you are pixel peeping, forget about it. With your 26 MP 80D on a typical 23" HD (1920 x 1024 screen pixels) monitor, viewing at 100% is the equivalent of enlarging to about 60x40 inches, an enlargement factor of approximately 100X! With the FF 26MP 6D Mark II, the same 100% view on the same monitor has an enlargement factor of about 64X. To look at it another way, if you are using a lens that can resolve 80 lines per millimeter with a full frame camera, you'd need a lens capable of 128 lpmm on a 1.6X crop sensor to get the same final resolution at the same final display size!
• The distinguishing factor for Canon's "L" lenses is not that they are necessarily sharper than their non-L counterparts. It's that they have a higher build quality, greater resistance to environmental factors such as dust and moisture, and are designed to be much more durable and able to take the constant pounding which professional photographers and the work that they do day in and day out will subject to their equipment. There's nothing "magical" about L glass. For instance, the EF-S 17-55mm f/2.8 IS is better optically from about 24mm and up on a crop sensor than the EF 16-35mm f/2.8L II. It's also nearly the optical equal of the original EF 24-70mm f/2.8L when the "L" is on a FF body and the 17-55mm is on one of the more recent crop bodies. The newer EF 24-70mm f/2.8 L II is a lot better optically from 24mm all the way to 70mm at every aperture, even the wider apertures, because the newer high resolution cameras revealed the flaws of the older 24-70mm. Ditto for the newer EF 16-35mm f/2.8 L III.
• The primary advantages of your two specific "L" lenses are in the areas of durability and resistance to adverse environmental conditions. Those attributes are critical to working pros who put their gear through the wringer every working day. Yes, they are a little better optically than some cheaper non-L zoom lenses. Their "sweet spots" may be larger in terms of the focal lengths and apertures at which they perform at a slightly higher level. But they are much closer to many non-L lenses in terms of optical performance than they are to the true premium zoom lenses in each category such as the EF 24-70mm f/2.8L II, or the EF 70-200mm f/2.8L IS III. Even those lenses only make a noticeable difference when used at or nearly at wide open apertures. At f/5.6 or f/8 there's very little real world difference in optical quality. At the f/5.6-f/8-f/11 apertures used for most landscape shooting pretty much every current lens from Canon, Nikon, Sony, Pentax, etc. are very good across the frame assuming they are in proper alignment and are designed to have a flat field. There are some very expensive lenses designed for certain uses that leave field curvature uncorrected. That's one of the things that makes the EF 85mm f/1.2L a "portrait lens" with such a unique look. Such a lens, even though costing thousands of dollars, is not a suitable lens for most landscape work.
• The real "more for less" value with regard to landscape lenses are in the "mid-grade" prime lens space. Since most landscape photography is done at relatively narrow apertures, one does not need to buy prime lenses with very wide apertures. As previously mentioned, it's also often the case that those very expensive wide aperture primes are designed for purposes other than what is optimal for landscape photography. A lens like the $550 EF 24mm f/2.8 IS will perform better at typical landscape apertures² from one edge of the frame to the other than the far more expensive $1,550 EF 24mm f/1.4 L II! At 24mm and f/5.6, the EF 24-105mm f/4 L is about as sharp as the EF 24mm f/1.4 L II, but not as sharp as the EF 24mm f/2.8 IS, though the zoom² does show about twice as much geometric distortion and vignettes more in the corners than either prime does. At f/5.6 the old 1990 vintage EF 35mm f/2 and the newer $550 EF 35mm f/2 IS that replaced it in 2012 are both the equal of the more expensive EF 35mm f/1.4 L, though the EF 35mm f/1.4 L II is slightly better (for $1,700 it ought to be!).

So should you go out tomorrow and swap your 80D + EF 24-105mm f/4 L IS and EF 70-200mm f/4 L for a 6D Mark II and an EF 35mm f/2 plus EF 85mm f/1.8 and either an EF 135mm f/2 L or EF 200mm f/2.8 L?

Probably not. Keep using what you have until you've reached the limits of their performance and can articulate exactly why those pieces of gear are limiting you. Then you'll know what kind of gear you need to use to overcome those limitations.

_{¹ Bear in mind that most of the amazing landscape photos you see at sites such as 500px and flickr have extensive post processing applied that tends to sharpen the results compared to how the image first looks straight out of the camera. That is not to say that getting a great exposure and composition in great light doesn't matter. It just means that many times a great capture is only the starting point of a great image. In some cases advanced techniques such as focus stacking and highly detailed lens correction are being applied. The 'Unsharp mask' tool can have a remarkable effect on the perceived sharpness of an image. So can the 'Clarity' (local contrast) slider. So can using an HSL/HSV/HSB tool to individually control the hue, saturation and luminance/value/brightness of different colors. Saving raw files and putting in the post-processing work is required to get everything out of a landscape shot, no matter what the equipment used to capture it.}

_{² To see the actual test data at DxO Mark, rather than the useless single number "effective megapixels" score, please click on 'Measurements' → 'Sharpness' → 'Profiles' and select the appropriate focal length (for zoom lenses) and aperture desired for each lens.}

Conclusion

In the end, gear with higher capabilities can certainly help. But a better camera or better lenses won't make you a better photographer. It will just allow you to use more of the skill, knowledge, and experience you've picked up along the way. Part of that experience and knowledge contributes to the ability to pick the best tool for the job from among the options one has available.

Answer 2

At f/11 on a high-megapixel crop sensor camera, the sharpness of the lens is reduced by diffraction. Prefer f/8 and wider. If you want maximum depth of field for pixel-peeping, f/8 should be your chosen aperture because narrower apertures have diffraction and wider apertures have lesser depth of field.

If you want the best possible picture, shoot RAW and use the Canon Digital Photo Professional and its Digital Lens Optimizer feature.

However, I don't understand why you would like to use a 24-105L lens on a crop sensor camera. You lack the 18-24mm range, and it's not much sharper either when used on a crop camera. The f/4 isn't much wider aperture than f/5.6 which the typical cheap crop lenses have on the tele end.

The same is true for 70-200 f/4L. It is not any sharper than the 55-250mm lens when used on a crop camera.

If using a crop sensor camera and not needing a fast aperture such as f/2.8, my choice would be the cheap STM-focused crop sensor specific lenses, or the 18-135 USM.

I wouldn't proceed with the plan to purchase L lenses for crop cameras. There may be justified reasons to use L lenses on crop cameras like 400 f/5.6L for daylight bird photography, or 70-200 f/2.8L for low light sports photography in situations where you would be using a 1.4x extender with a full frame camera anyway (so the inherent 1.6x extender light loss present in crop cameras doesn't matter), but landscape photography is not one of them.

If you have the money to spend on lens/lenses, I would choose the Canon 18-135mm USM lens. Marvelous image stabilization, fast autofocus and effectively two lenses (standard + tele) in one. The way they made this possible is narrow aperture, but in daylight landscape photography you won't be using wide apertures anyway so it shouldn't matter.

If you don't manually clean the sensor regularly, a 18-135 USM may have the best image quality of all possible lenses, because you never need to change lenses if the only type of photography you do is landscape photography, resulting in less sensor dust. At narrow apertures like the ones sometimes used for landscape photography, sensor dust may become an issue. Sensor dust will ruin your image quality.

Further increasing the practical image quality of the Canon 18-135mm lens is its light weight. The worst image is the image you never take because the camera and all the lenses are too heavy to carry with you. As they say, the best camera is the one you have with you.

Spend the rest on a monopod, a monopod tilt head, a tripod (+ head) and a tripod carrying bag.